Author_Institution :
Appl. Phys. Lab., Johns Hopkins Univ., Laurel, MD
Abstract :
A large scale network simultaneously experiences multiple and varied events that degrade its ability to serve its customers. Network management of large networks that span multiple networks, network technologies, network events (faults, performance, security, other disruptions) is a multidimensional problem. While, policy based network management frameworks have been proposed to help manage the complexity, additional frameworks are needed to help manage network of networks. In this paper, a finite state machine based aggregation of network, events, N-State, is proposed, as a framework to manage complex networks and as a PBNM framework extension to manage network of networks. N-state, targeted network behavior for each state, and collaboratively developed PBNM policies mapped to state and targeted behavior, provide a means for network operators to comprehend overall network situation, assess overall network degradation, collaborate in restoration of network services and intervene when necessary
Keywords :
computer network management; finite state machines; N-state driven policy-based network management; PBNM framework; complex network management; end-end network behavior control; finite state machine based aggregation; large scale network; multiple networks; network events; network services; overall network degradation; Automata; Automation; Collaboration; Complex networks; Degradation; Laboratories; Large-scale systems; Physics; Switches; Technology management;